Explore the vast range of titles available.

Popular frameworks for studying habitat selection include resource‐selection functions (RSFs) and step‐selection functions (SSFs), estimated using logistic and conditional logistic regression, respectively. Both frameworks compare environmental covariates associated with locations animals visit with environmental covariates at a set of locations assumed available to the animals. Conceptually, slopes that vary by individual, that is, random coefficient models, could be used to accommodate inter‐individual heterogeneity with either approach. While fitting such models for RSFs is possible with standard software for generalized linear mixed‐effects models (GLMMs), straightforward and efficient one‐step procedures for fitting SSFs with random coefficients are currently lacking. To close this gap, we take advantage of the fact that the conditional logistic regression model (i.e. the SSF) is likelihood‐equivalent to a Poisson model with stratum‐specific fixed intercepts. By interpreting the intercepts as a random effect with a large (fixed) variance, inference for random‐slope models becomes feasible with standard Bayesian techniques, or with frequentist methods that allow one to fix the variance of a random effect. We compare this approach to other commonly applied alternatives, including models without random slopes and mixed conditional regression models fit using a two‐step algorithm. Using data from mountain goats (Oreamnos americanus) and Eurasian otters (Lutra lutra), we illustrate that our models lead to valid and feasible inference. In addition, we conduct a simulation study to compare different estimation approaches for SSFs and to demonstrate the importance of including individual‐specific slopes when estimating individual‐ and population‐level habitat‐selection parameters. By providing coded examples using integrated nested Laplace approximations (INLA) and Template Model Builder (TMB) for Bayesian and frequentist analysis via the R packages R‐INLA and glmmTMB, we hope to make efficient estimation of RSFs and SSFs with random effects accessible to anyone in the field. SSFs with individual‐specific coefficients are particularly attractive since they can provide insights into movement and habitat‐selection processes at fine‐spatial and temporal scales, but these models had previously been very challenging to fit. The authors provide a coherent framework for fitting resource‐selection functions (RSFs) and step‐selection functions (SSFs) with random effects. To allow fitting of SSFs, the authors reformulate the conditional logistic regression model as a (likelihood‐equivalent) Poisson model, where stratum‐specific intercepts are included as a random effect with a fixed large prior variance.

Advances in tracking technology have led to an exponential increase in animal location data, greatly enhancing our ability to address interesting questions in movement ecology, but also presenting new challenges related to data management and analysis. Step‐selection functions (SSFs) are commonly used to link environmental covariates to animal location data collected at fine temporal resolution. SSFs are estimated by comparing observed steps connecting successive animal locations to random steps, using a likelihood equivalent of a Cox proportional hazards model. By using common statistical distributions to model step length and turn angle distributions, and including habitat‐ and movement‐related covariates (functions of distances between points, angular deviations), it is possible to make inference regarding habitat selection and movement processes or to control one process while investigating the other. The fitted model can also be used to estimate utilization distributions and mechanistic home ranges. Here, we present the R package amt (animal movement tools) that allows users to fit SSFs to data and to simulate space use of animals from fitted models. The amt package also provides tools for managing telemetry data. Using fisher (Pekania pennanti) data as a case study, we illustrate a four‐step approach to the analysis of animal movement data, consisting of data management, exploratory data analysis, fitting of models, and simulating from fitted models. New tracking technologies allow users to collect large amount of data and address entirely new questions. The amt (animal movement tools) R package provides tools to manage telemetry data and to fit step‐selection functions and resource‐selection functions.

Habitat‐selection analyses are often used to link environmental covariates, measured within some spatial domain of assumed availability, to animal location data that are assumed to be independent. Step‐selection functions (SSFs) relax this independence assumption, by using a conditional model that explicitly acknowledges the spatiotemporal dynamics of the availability domain and hence the temporal dependence among successive locations. However, it is not clear how to produce an SSF‐based map of the expected utilization distribution. Here, we used SSFs to analyze virtual animal movement data generated at a fine spatiotemporal scale and then rarefied to emulate realistic telemetry data. We then compared two different approaches for generating maps from the estimated regression coefficients. First, we considered a naïve approach that used the coefficients as if they were obtained by fitting an unconditional model. Second, we explored a simulation‐based approach, where maps were generated using stochastic simulations of the parameterized step‐selection process. We found that the simulation‐based approach always outperformed the naïve mapping approach and that the latter overestimated home‐range size and underestimated local space‐use variability. Differences between the approaches were greatest for complex landscapes and high sampling rates, suggesting that the simulation‐based approach, despite its added complexity, is likely to offer significant advantages when applying SSFs to real data.

For gregarious animals the cost-benefit trade-offs that drive habitat selection may vary dynamically with group size, which plays an important role in foraging and predator avoidance strategies. We examined how habitat selection by bison (Bison bison) varied as a function of group size and interpreted these patterns by testing whether habitat selection was more strongly driven by the competing demands of forage intake vs. predator avoidance behavior. We developed an analytical framework that integrated group size into resource selection functions (RSFs). These group-size-dependent RSFs were based on a matched case-control design and were estimated using conditional logistic regression (mixed and population-averaged models). Fitting RSF models to bison revealed that bison groups responded to multiple aspects of landscape heterogeneity and that selection varied seasonally and as a function of group size. For example, roads were selected in summer, but not in winter. Bison groups avoided areas of high snow water equivalent in winter. They selected areas composed of a large proportion of meadow area within a 700-m radius, and within those areas, bison selected meadows. Importantly, the strength of selection for meadows varied as a function of group size, with stronger selection being observed in larger groups. Hence the bison-habitat relationship depended in part on the dynamics of group formation and division. Group formation was most likely in meadows. In contrast, risk of group fission increased when bison moved into the forest and was higher during the time of day when movements are generally longer and more variable among individuals. We also found that stronger selection for meadows by large rather than small bison groups was caused by longer residence time in individual meadows by larger groups and that departure from meadows appears unlikely to result from a depression in food intake rate. These group-size-dependent patterns were consistent with the hypothesis that avoidance of predation risk is the strongest driver of habitat selection.

Although classical ecological theory (e.g., on ideal free consumers) recognizes the potential effect of population density on the spatial distribution of animals, empirical species distribution models assume that species-habitat relationships remain unchanged across a range of population sizes. Conversely, even though ecological models and experiments have demonstrated the importance of spatial heterogeneity for the rate of population change, we still have no practical method for making the connection between the makeup of real environments, the expected distribution and fitness of their occupants, and the long-term implications of fitness for population growth. Here, we synthesize several conceptual advances into a mathematical framework using a measure of fitness to link habitat availability/selection to (density-dependent) population growth in mobile animal species. A key feature of this approach is that it distinguishes between apparent habitat suitability and the true, underlying contribution of a habitat to fitness, allowing the statistical coefficients of both to be estimated from multiple observation instances of the species in different environments and stages of numerical growth. Hence, it leverages data from both historical population time series and snapshots of species distribution to predict population performance under environmental change. We propose this framework as a foundation for building more realistic connections between a population's use of space and its subsequent dynamics (and hence a contribution to the ongoing efforts to estimate a species' critical habitat and fundamental niche). We therefore detail its associated definitions and simplifying assumptions, because they point to the framework's future extensions. We show how the model can be fit to data on species distributions and population dynamics, using standard statistical methods, and we illustrate its application with an individual-based simulation. When contrasted with nonspatial population models, our approach is better at fitting and predicting population growth rates and carrying capacities. Our approach can be generalized to include a diverse range of biological considerations. We discuss these possible extensions and applications to real data.

Understanding how animals use their habitat is essential to understand their biology and support conservation efforts. Technological advances in tracking technologies allow us to follow animals at increasingly fine temporal resolutions. Yet, how tracking devices' sampling intervals impact results remains unclear, as well as which method to use. Using simulations and empirical data from wild boars tracked in Germany, we systematically examine how the temporal resolution of movement data in interaction with the spatial autocorrelation of the landscape affects the outcomes of two common techniques for analyzing habitat selection: resource‐selection analysis (RSA) and an autocorrelation‐informed weighted derivative (wRSA) as well as integrated step‐selection analysis (iSSA). Each method differs in the definition of “available” locations (RSA) and the implementation of the movement model during parameter estimation (iSSA). Our simulations suggested that landscape autocorrelation has a much stronger effect on the estimated selection coefficients and their variability than the sampling interval. Higher sampling intervals (i.e., longer time between steps) are required for landscapes with high autocorrelation, enabling the animal to experience enough variability in clumped landscapes. Short sampling intervals generally lead to higher variability and fewer statistically significant estimates (in particular for wRSA). Our results complement recent attempts to outline a coherent framework for habitat‐selection analyses and to explain them to practitioners. We further contribute to these efforts by assessing the sensitivity of two commonly used methods, RSA and iSSA, to the changes in sampling interval of movement data. We expect our findings to further raise awareness of pitfalls underlying the comparison of estimated selection coefficients obtained in different studies and to assist movement ecologists in choosing the appropriate method for habitat‐selection analysis. Advances in tracking technology enable finer temporal resolution in animal movement data, but the impact of sampling intervals on habitat‐selection analysis remains unclear. Using simulations and wild boar data from Germany, researchers found that higher sampling intervals are needed in highly autocorrelated landscapes to capture habitat variability, while shorter intervals result in greater variability and fewer significant estimates. These findings contribute to a broader framework for improving habitat‐selection analyses and guiding practitioners.

The resource hierarchy hypothesis predicts that the most important factors limiting a species’ distribution act at the coarsest spatial scales. However, resource selection behaviour affords mobile organisms the opportunity to adopt a range of tactics for navigating spatial trade‐offs between competing biotic and abiotic constraints. Throughout the animal kingdom, partial migration (where some individuals migrate, and others remain resident year round) offers a pervasive example of such behavioural polymorphism. Identifying the differences between these behaviours is therefore central to understanding the conditions (habitat) needed to sustain migrant and resident populations. Here we test an extension of the resource hierarchy hypothesis. We hypothesized that rather than responding to a single limiting factor, migration and residency represent contrasting scale‐specific approaches to managing trade‐offs between forage, predation risk and severe winter conditions. Furthermore, we predicted that the distribution of habitat selected by migrants and residents is predictive of the local prevalence of migratory behaviour. To test these hypotheses, we quantified migratory status‐ (resident/migrant) and season‐specific (winter/summer) differences in resource selection by eight populations of federally endangered Sierra Nevada bighorn sheep Ovis canadensis sierrae across three spatial scales: population range, individual range and within individual range. We then integrated these spatial predictions to produce separate spatial predictions of migrant and resident winter habitat. As predicted, model selection provided strong evidence for the importance of status‐specific differences in resource selection. Residents showed stronger coarse‐scale selection for terrain associated with predator avoidance and stronger fine‐scale selection for greenness, while in migrants this pattern was reversed. Availability of migrant habitat predicted the local prevalence of migration (top model pseudo R2 of .87). Our ability to respond to global declines of migratory species depends on improving our understanding of the conditions required to maintain migratory behaviour. Through explicitly contrasting migrant and resident behaviour, our results illustrate seasonal differences in migrant and resident habitat and how these two behaviours represent responses to different limiting conditions. Our analyses provides a novel empirical basis for assessing the local prevalence of migratory behaviour across large landscapes. The authors present a novel approach to identifying seasonal habitat for partially migratory species. Through a case study of an endangered ungulate, they demonstrate that behavioral models accounting for migratory tactics can predict the local prevalence of migratory behavior. These methods have great potential to inform conservation priorities for migratory species.

We review 87 articles published in the Journal of Wildlife Management from 2000 to 2004 to assess the current state of practice in the design and analysis of resource selection studies. Articles were classified into 4 study designs. In design 1, data are collected at the population level because individual animals are not identified. Individual animal selection may be assessed in designs 2 and 3. In design 2, use by each animal is recorded, but availability (or nonuse) is measured only at the population level. Use and availability (or unused) are measured for each animal in design 3. In design 4, resource use is measured multiple times for each animal, and availability (or nonuse) is measured for each use location. Thus, use and availability measures are paired for each use in design 4. The 4 study designs were used about equally in the articles reviewed. The most commonly used statistical analyses were logistic regression (40%) and compositional analysis (25%). We illustrate 4 problem areas in resource selection analyses: pooling of relocation data across animals with differing numbers of relocations, analyzing paired data as though they were independent, tests that do not control experiment wise error rates, and modeling observations as if they were independent when temporal or spatial correlations occurs in the data. Statistical models that allow for variation in individual animal selection rather than pooling are recommended to improve error estimation in population-level selection. Some researchers did not select appropriate statistical analyses for paired data, or their analyses were not well described. Researchers using one-resource-at-a-time procedures often did not control the experiment wise error rate, so simultaneous inference procedures and multivariate assessments of selection are suggested. The time interval between animal relocations was often relatively short, but existing analyses for temporally or spatially correlated data were not used. For studies that used logistic regression, we identified the data type employed: single sample, case control (used–unused), use–availability, or paired use–availability. It was not always clear whether studies intended to compare use to nonuse or use to availability. Despite the popularity of compositional analysis, we do not recommend it for multiple relocation data when use of one or more resources is low. We illustrate that resource selection models are part of a broader collection of statistical models called weighted distributions and recommend some promising areas for future development.

Applications of logistic regression in a used–unused design in wildlife habitat studies often suffer from asymmetry of errors: used resource units (landscape locations) are known with certainty, whereas unused resource units might be observed to be used with greater sampling intensity. More appropriate might be to use logistic regression to estimate a resource selection function (RSF) tied to a use–availability design based on independent samples drawn from used and available resource units. We review the theoretical motivation for RSFs and show that sample “contamination” and the exponential form commonly assumed for the RSF are not concerns, contrary to recent statements by Keating and Cherry (2004; Use and interpretation of logistic regression in habitat-selection studies. Journal of Wildlife Management 68:774–789). To do this, we re-derive the use–availability likelihood and show that it can be maximized by logistic regression software. We then consider 2 case studies that illustrate our findings. For our first case study, we fit both RSFs and resource selection probability functions (RSPF) to point count data for 4 bird species with varying levels of occurrence among sample blocks. Drawing on our new derivation of the likelihood, we sample available resource units with replacement and assume overlapping distributions of used and available resource units. Irrespective of overlap, we observed approximate proportionality between predictions of a RSF and RSPF. For our second case study, we evaluate the classic use-availability design suggested by Manly et al. (2002), where availability is sampled without replacement, and we systematically introduce contamination to a sample of available units applied to RSFs for woodland caribou (Rangifer tarandus caribou). Although contamination appeared to reduce the magnitude of one RSF beta coefficient, change in magnitude exceeded sampling variation only when >20% of the available units were confirmed caribou use locations (i.e., contaminated). These empirically based simulations suggest that previously recommended sampling designs are robust to contamination. We conclude with a new validation method for evaluating predictive performance of a RSF and for assessing if the model deviates from being proportional to the probability of use of a resource unit.